Why RTK Wasn't Enough (And What I Added)

RTK (Reduce Toolkit) is a solid Rust CLI for reducing AI context size. I used it daily. Then I hit its limits. RTK handles ANSI stripping and basic deduplication well. But real-world CLI output has patterns RTK doesn't catch. I forked it and built ContextZip. What RTK Does Well ANSI escape code removal Basic line deduplication Character count reporting Clean Rust codebase What Was Missing 1. Language-aware stack trace filtering. RTK treats stack traces as plain text. It doesn't know that node:internal/modules/cjs/loader is a framework frame and /app/src/server.ts is your code. ContextZip recognizes stack trace formats for Node.js, Python, Rust, Go, Java, and C# — and strips framework frames while keeping application frames. 2. Semantic duplicate grouping. RTK deduplicates exact matches. But 40 TypeScript errors with the same message but different file paths aren't exact duplicates. ContextZip groups by error pattern, not exact string match. 3. Command-specific patterns. npm install pro